Sentinel is a local, autonomous AI agent designed to keep Master Software Engineers and Architects "in the know."
Unlike standard RSS feeds or newsletters, Sentinel uses LangGraph to build a cyclic agentic workflow that reads, filters, analyzes, and synthesizes high-signal intelligence from ArXiv, GitHub, and Hacker News.
It is specifically tuned to filter noise and prioritize:
- Generative AI & LLM Architectures (RAG, Agents)
- Enterprise Software Engineering (.NET, Microservices)
- LegalTech & Digital Forensics
- 🧠 Semantic Gatekeeper: Uses LLMs (Claude 3.5 Sonnet or Gemini 1.5 Pro) to grade papers on a 1-10 scale based on a custom "Interest Constitution."
- 🕵️ Code Hunter: Automatically scrapes ArXiv abstract pages to find and verify official GitHub implementations.
- 📡 Multi-Source Monitoring:
- ArXiv: Scans
cs.SE,cs.AI,cs.CR,cs.CL. - GitHub Trending: Monitors C# and Python repos.
- Hacker News: Filters for high-impact tech discussions.
- ArXiv: Scans
- 💾 Long-Term Memory: Uses SQLite to deduplicate findings (never reads the same paper twice).
- 📰 Daily Briefing: Generates a clean, Obsidian-friendly Markdown report in
Daily_Briefings/.
The system runs on LangGraph with the following cyclic node structure:
Monitor (Ingestion) -> Industry_Monitor (Git/HN) -> Gatekeeper (LLM Filter) -> Analyst (Code Search) -> Journalist (Report Gen)
git clone [https://github.com/abahjat/sentinel.git](https://github.com/abahjat/sentinel.git)
cd sentinel